Kev's Defensive Breakdown: Lakers at Warriors (March 15, 2010)

by duaneok66 on Mar 17, 2010 12:51 PM PDT

Note *** If you wonder what this is all about – it’s all explained in my original fanpost “Kev’s defensive breakdown: Raptors at Warriors (March 13, 2010)” ***

The team score came up (-5). Compare that to a score of zero – which is worse than the game against the Raptors, but that is to be expected due to the competition. (Still, the Thunder scored a +28 last time they faced the Lakers). The team seemed more active versus the Lakers (and maybe because the game was on ESPN). Still, it’s tough to guard the Lakers because they have THREE guys that truly command a double team (Kobe Bryant, Pau Gasol, and Andrew Bynum). Most teams in the NBA have no more than one. So it was a decent showing considering all the circumstances (like the injuries).

C J Watson (+10)

Great game for C J. He managed to get three steals in less than twenty-four minutes of game time. His only real “mistake” was giving up two points to Kobe Bryant in the post, but that’s to be expected when you’re giving up four inches and thirty pounds.

Reggie Williams (+2)

He only played four minutes, but he contested a Ron Artest shot and picked up a loose ball in that time frame – no mistakes.

Devean George (zero)

Anthony Morrow (zero)

I don’t think that Morrow thought he’d be guarding guys like Lamar Odom (6-10, 220) when he made it to the NBA. Still, my numbers show his only miscue was a blown block out. That led to a Lamar Odom putback in the second quarter.

Anthony Tolliver (-1)

It was a decent game considering he was playing against the twin towers all night. I would have guessed his score would have been -10 before the game started, but he managed to get two steals and a two deflections to help offset the inside scoring.

Stephen Curry (-2)

Since most of the action was down low, Curry had a pretty quiet night guarding fossil Derek Fisher. He should get a better test tonight against Darren Collson. Still that foul where he gave up three free throws to Derek Fisher was a no-no.

Chris Hunter (-4)

Corey Maggette (-4)

He (like Monta) also has a tendency to lose focus on the defensive side of the ball. Early in the third quarter, Corey is guarding Ron Artest. Kobe comes to set a downpick to free Artest uptop. The ball is on the other side of the court, and Corey is locked in on it – Artest is at the top of the key before Corey spots him. Kobe’s pick wasn’t even really necessary! Artest drains the two easily.

Monta Ellis (-6)

The good:

· Stripped Kobe Bryant (1st - 10:20)

· Contested Bryant (1st - 9:30)

· Deflection (1st – 3:40)

· Forced Turnover deflected it off Shannon Brown out of bounds (2nd – 11:00)

· Contested Jordan Farmar (2nd – 10:00)

· Contested Kobe Bryant (2nd – 3:10)

· Loose ball steal from Bynum (2nd – 2:30)

· Drew offensive foul from Bryant (2nd – 1:00)

· Steal from Kobe Bryant (End of 1st half)

· Deflection (3rd – 3:35)

· Loose Ball (3rd – 3:30)

· Contested Bryant (4th – 6:30)

· Defensed Bryant in the post (4th – 5:40)

· Contested Derek Fisher (4th – 1:30)

· Stripped Bryant (4th – 0:30)

And the Bad:

· needed double team (DT) on Bryant – Gasol hits a jumper (1st – 9:40)

· lazy transition defense (1st – 9:30)

· posted by Bryanr (1st – 4:40)

· needed double team – Ron Artest for three (1st – 3:20)

· bad transition defense (1st – 3:00)

· posted by Kobe Bryant (1st – 0:45)

· dumb swipe at the ball (lazy defense), gave up defensive position; forced Chris Hunter to rotate and help; shot was missed but there was a free lane for an offensive rebound; Anthony Morrow was forced to foul Pau Gasol as a result (This play is my rebuttal for when announcers cry “No one blocked out.” Well, it’s hard to block out and help out on driving guards at the same time. The fault lies on the perimeter – in this case with Monta) (End of first quarter)

· Needed double team help; Odom scores a layup after defensive rotation (mid second quarter)

· Bad transition defense (2nd 7:00)

· Failure to contest a Kobe jumper (2nd 5:00)

· Gave up penetration to Kobe – shooting foul (2nd – 3:40)

· Shooting Foul on Bryant (3rd 8:40)

· Gave up Penetration to Bryant (3rd 5:40)

· Shooting Foul on Bryant (3rd - :40)

· Posted by Lamar Odom (4th – 9:50)

· Gave up Penetration to Bryant; Hunter rotated to help and fouled (4th – 6:10)

What is glaring on that list are the three mistakes in transition – my spreadsheet showed that no one else had any. Let’s look at one of them. In the middle of the second quarter, C J Watson races to the basket and misses a layup. Lakers guard Shannon Brown gets the miss and heads upcourt. Monta is right alongside of him. In my opinion (remember this analysis is somewhat objective), Monta should have raced back with Brown. Instead, he takes the easy way out. He comes up from behind and tries to swipe the ball out. In doing so, he gives up a step or two and loses momentum. He misses, and Brown is off. The Lakers scored easily on an Odom layup. That’s one of my complaints with his defensive play – he’d rather get a steal and get the highlight play versus playing sound defense and getting back. The former requires much less effort.

This FanPost is a submission from a member of the mighty Golden State of Mind community. While we're all here to throw up that W, these words do not necessarily reflect the views of the GSoM Crew. Still, chances are the preceding post is Unstoppable Baby!

Tweet Comment 67 comments | 5 recs |

Do you like this story?

Comments

Question

How does Kobe posting up qualify as bad? What did he do in the post? Did he stop him or did he go by him? The good should include how he worked Kobe the whole game. Also I thought Steph did very well on defense.

We all love the Warriors..
We all hate Cohan as an owner of this franchise…

by dubzfan on Mar 17, 2010 2:04 PM PDT reply actions

He only counts it when the other team scored, so “posted up by Kobe” means Kobe scored on it, I think.

by Missing Barry on Mar 17, 2010 2:23 PM PDT up reply actions

correct

by duaneok66 on Mar 17, 2010 3:12 PM PDT up reply actions

great info and effort

by dubtown on Mar 17, 2010 2:17 PM PDT reply actions

Rec again! It’s really great seeing this stuff broken down and systematically counted and recorded. So it appears to me that Monta caused 6 turnovers based on your judgments, correct?

by Missing Barry on Mar 17, 2010 2:25 PM PDT reply actions

that is correct!

by duaneok66 on Mar 17, 2010 3:17 PM PDT up reply actions

I appreciate the effort here, but the results here show (maybe just in my opinion) some flaws with this system.

By most all accounts, Monta played pretty solid D against Kobe in this game (writers, Nellie, etc.) Also, his D for the most part passed the eye test as well (it looked like he was frustrating Kobe and Kobe had, for him, a pretty bad game).

- “Needed a double team” doesn’t work a lot of the time, as the double team might not be necessary, but one of the others could do it anyway. Say Odom has the ball in the post and is being defended OK by Tolliver, but Monta decides to leave Kobe anyway, and Kobe ends up hitting a jumper. This shouldn’t go against Tolliver.

- A shooting foul on a breakaway on a poor FT shooter to prevent the layup may actually be a positive.

- If somebody else misses an assignment and one of the guards ends up guarding a skilled big (like Gasol) very likely they are going to get posted up, and probably give up a basket. I wouldn’t consider that bad on that player necessarily.

Again, this type of analysis is going in the right direction, but has its flaws.

by FitzPRSpin on Mar 17, 2010 2:43 PM PDT reply actions

By most all accounts, Monta played pretty solid D against Kobe in this game (writers, Nellie, etc.) Also, his D for the most part passed the eye test as well (it looked like he was frustrating Kobe and Kobe had, for him, a pretty bad game).

Honestly, I’m not sure this is a valid criticism. One the one hand, you have writers who know nothing about basketball, and Nellie (who knows far more than any of us, but is Nellie, who can ever really know his motivation for anything?), and on the other hand, you have someone who systematically watched and categorized the game and recorded every play Monta was involved in defensively. There might be some problems with methodology, but you can plainly see plenty of instances of bad D the author listed on Monta’s part. It really might be that his defense was a net negative.

by Missing Barry on Mar 17, 2010 2:56 PM PDT up reply actions

I guess my problem with “seeing plenty of instances of bad D” is that close to half of these instances are purely subjective (lazy transition D, bad transition D, “dumb swipe at the ball”), not necessarily true (needed a double team vs. a double team happened, needed or not) or a situation which wasn’t necessarily bad D on his part (posted up by Odom probably should lead to a basket on 6’3" Monta).

The concept of this type of analysis is good, just appears to need a lot of tweaks.

by FitzPRSpin on Mar 17, 2010 3:13 PM PDT up reply actions

And I’m not claiming to have a solution, but it was pretty evident watching, and stats wise, that the front line got owned by the Lakers (28/41 FG 20 FT given up, 12 Offensive boards given up, looking generally bad) but this doesnt show up in their +/- rating here.

Plus, not to nitpick, but Monta had 5 steals, plus apparently a drawn charge and a ball knocked out of bounds, which is more than the 6 TO cause came up with here.

by FitzPRSpin on Mar 17, 2010 3:20 PM PDT up reply actions

I do penalize post players for not blocking out (if it leads to a basket AND penetration wasnt given up which causes bigs to rotate) . . .

I was originally going to penalize perimeter players for not boxing out – but after charting several games I diescovred that those players RARELY block out in the NBA . . .

by duaneok66 on Mar 17, 2010 3:34 PM PDT up reply actions

Just another note on the scoring system, a player gets a negative for both going under a screen on a 3pt shooter and also giving up penetration.

Say he goes over the screen, which causes the shooter to penetrate instead. This seems to always give a negative on a screen if the player scores, whether or not “poor defense” occurred.

by FitzPRSpin on Mar 17, 2010 3:25 PM PDT reply actions

A player is only penalized for giving up penetration when there is no pick.

As to your other concerns – I now it is subjective and crude, I amdmitted that upfront in my other post. Still, Kobe got 29 points on 18 shots. Even throwing in the turnovers, that’s not a bad outing. It’s not Monta’s fault (6-3 guarding 6-6) but when you have to send double teams the whole game, you really can’t say that “Monta did a marvelous job on him” What else is Nellie going to say. And Monta is lazy in transition – that’s why I included the times in case people want to check it out themselves.

Thanks for the feedback . . .

by duaneok66 on Mar 17, 2010 3:32 PM PDT up reply actions

Got ya. Thanks for the clarification.

I’ll just respectfully agree to disagree on this being a good outing for Kobe. He only took 18 shots to get the 29, but turning it over 9 times mostly against a smaller defender and getting an assist and 2 boards less than normal doesn’t qualify as a good outing in my book.

I’ll agree that Nellie’s Defensive praises don’t mean much, but watching the game it sure appeared to me that the D on Kobe was considerably better than anything the front court was doing. Maybe add a stat for “looking completely overmatched” in the case of Gasol/Bynum/Odom vs. Hunter/Tolliver/Maggete.

by FitzPRSpin on Mar 17, 2010 3:42 PM PDT up reply actions

good points . . . actually Monta’s “defensive usage” (for lack of a better term) was the highest for one single game for any player Ive tracked all season (Ive scored almost all the Thunder games this year). So his -6 was not that bad considering.

Seriously, thanks for the comments – I appreciate the intelligent back and forth . . .

by duaneok66 on Mar 17, 2010 3:47 PM PDT up reply actions

Someone’s head is going to explode if this morphs into a adjusted defensive +/- based on usage stat.

by FitzPRSpin on Mar 17, 2010 3:51 PM PDT up reply actions

Tonight's game

There seems to be enough interest, so I will post a breakdown of tonight’s game tomorrow morning sometime . . .

by duaneok66 on Mar 17, 2010 3:53 PM PDT reply actions

Thanks, man. I love this stuff! Great analysis.

Welcome to the Pit of Despair! Don't even think about trying to escape.

by Naticus2 on Mar 17, 2010 4:00 PM PDT up reply actions

thanks!

much appreciated . . .

by duaneok66 on Mar 17, 2010 4:05 PM PDT up reply actions

I have to say, a lot of the blame goes to Don Nelson for playing Monta too many minutes. On some level, he must be trying to conserve energy at times, having played 40+ minutes a night every night practically.

Welcome to the Pit of Despair! Don't even think about trying to escape.

by Naticus2 on Mar 17, 2010 4:01 PM PDT reply actions

continual refinement

I think you’re on to something, Duane. And, I agree with Fitz that the point system is missing some important criteria. I might add that double-teams are often employed as an official strategy. Clearly, a coach would employ the “jordan rules” on a player such as Kobe. Maybe you account for this, I’m not sure. Did you ever work these points for premier defenders like Artest, Battier or maybe Barnes?

Could be an interesting community project.

"That’s Chavey; he’s a good athlete. He can play anywhere … except second base. He’s not that good." -M. Ellis

by eshock on Mar 17, 2010 4:56 PM PDT reply actions

Maybe you account for this, I’m not sure.

He does account for it. It’s called “sample size.” You don’t qualify Kobe’s offensive statistics by saying “Yeah, but he was going up against Bruce Bowen in his prime!”, do you? No, you just say “Yeah, I’d expect Kobe’s stats to suffer tonight because he’s going up against Bruce Bowen.”

I’d be interested to learn how these statistical measures play out over the long run.

Trust me, learning english isn’t a waste of time. It is actually sort of useful.

-randolphforpresident

by Dubs fan in Boston on Mar 17, 2010 5:36 PM PDT up reply actions

Being critical of Monta's D that night makes no sense

He hounded Kobe all night long. If you didn’t enjoy watching him scrap I don’t know what to tell you. Getting scored on by Kobe Bryant does not mean your D was bad. Last time I checked, he was pretty good at scoring the ball. Kobe was clearly frustrated all night and tied his career highs in turnovers. If you can’t appreciate the amount of effort that went into that, then I don’t know why you’re watching.

I mean can you honestly sit there and say Steph played better on defense that night guarding Fisher???!! He couldn’t even stay out of foul trouble guarding a guy who has been washed up for a while now.

Also, you said objective went you should of said subjective, which means it’s your opinion.

by sjsnider on Mar 17, 2010 6:25 PM PDT reply actions

He hounded Kobe all night long. If you didn’t enjoy watching him scrap I don’t know what to tell you. Getting scored on by Kobe Bryant does not mean your D was bad. Last time I checked, he was pretty good at scoring the ball. Kobe was clearly frustrated all night and tied his career highs in turnovers. If you can’t appreciate the amount of effort that went into that, then I don’t know why you’re watching.

Honestly, the author put a lot more thought into his opinions – actually objectively recording what happened on every single possession and scoring it, than you did with your opinion. I’m sure there are valid concerns with what he did, but simply criticizing the author because his opinion doesn’t agree with yours is not one of them. He watched and systematically recorded every single play that happened in the game. You did not. Based on that, his evaluation is more objectively and fact based than yours.

by Missing Barry on Mar 17, 2010 9:56 PM PDT up reply actions

He didn’t say Steph played better. He said that Steph scored better by a reasonably objective measure. Scoring more and playing better aren’t always synonymous.

Welcome to the Pit of Despair! Don't even think about trying to escape.

by Naticus2 on Mar 17, 2010 10:26 PM PDT up reply actions

Scoring more and playing better aren’t always synonymous.

As evidenced by Monta Ellis 09/10 vs. Monta Ellis 07/08…

Trust me, learning english isn’t a waste of time. It is actually sort of useful.

-randolphforpresident

by Dubs fan in Boston on Mar 18, 2010 10:43 AM PDT up reply actions

Just something im noticing in the Hornets game – Maggette is playing awful D on West and West is getting whatever he wants. However, none of the negatives in your system incorporate this (general lazy/bad play).

by FitzPRSpin on Mar 17, 2010 8:02 PM PDT reply actions

I think he does account for that. If West pots up and scores, that’s a minus. If West shoots and Maggette doesn’t contest, that’s a minus. If Maggette doesn’t keep up for transition defense, that’s a minus. I think he’s accounting for it. I predict Maggette, if he played bad defense, will not get a good score.

Welcome to the Pit of Despair! Don't even think about trying to escape.

by Naticus2 on Mar 17, 2010 10:28 PM PDT up reply actions

I’ll reserve judgement until I see the numbers. The reason I don’t think it will happen is in the Lakers game, Hunter, Tolliver, and Maggette all got absolutely worked by the front line but had reasonable numbers (-1, -4, -4). We will see though.

by FitzPRSpin on Mar 17, 2010 10:32 PM PDT up reply actions

Maggette's sore last night

. . . was bad (-7). I will have the breakdown up later than I thought (sorry) – it will be up sometime this afternoon . . .

by duaneok66 on Mar 18, 2010 7:50 AM PDT up reply actions

**** score (sorry for the typo)

by duaneok66 on Mar 18, 2010 7:56 AM PDT up reply actions

Maggette was a -6 less than four minutes into the game

details coming this afternoon . . .

by duaneok66 on Mar 18, 2010 7:53 AM PDT up reply actions

Keep em comming

These are great bro :)

by RUN EMC on Mar 17, 2010 11:24 PM PDT reply actions

thanks!

by duaneok66 on Mar 18, 2010 7:56 AM PDT up reply actions

I'm sorry

But any system that looks at the game on Monday and concludes that CJ Watson was the best defender of the game for the Warriors and Monta Ellis was the worst is a system that just does not work.

CJ Watson should win WDWY this year. Yeah he gets steals, because he goes for it EVERY TIME!

Did it count against Monta when Kobe shoved him out of bounds with his left hand (full elbow extension) and layed it in? If a player other than Kobe had done that it would have been called a foul and thus would have been a positive for Monta.

by WheresMyChippy on Mar 18, 2010 9:28 AM PDT reply actions

But any system that looks at the game on Monday and concludes that CJ Watson was the best defender of the game for the Warriors and Monta Ellis was the worst is a system that just does not work.

That’s a completely and utterly BS way to evaluate something. “I have an opinion based on no evidence, but someone taking the time to do research and come up with evidence is doing it wrong because it doesn’t agree with my opinion”. We need to ditch this mentality. It’s stupid. It doesn’t make sense.

Did it count against Monta when Kobe shoved him out of bounds with his left hand (full elbow extension) and layed it in? If a player other than Kobe had done that it would have been called a foul and thus would have been a positive for Monta.

Did Kobe score? Did it help the Lakers and hurt the Warriors? You mean, someone is actually being rated negative for a play that negatively affected our team? I don’t see what the objection is.

by Missing Barry on Mar 18, 2010 9:42 AM PDT up reply actions

Well put

That’s a completely and utterly BS way to evaluate something. "I have an opinion based on no evidence, but someone taking the time to do research and come up with evidence is doing it wrong because it doesn’t agree with my opinion". We need to ditch this mentality. It’s stupid. It doesn’t make sense.

I’d be curious to chart how often and under what conditions this syndrome is a permanent condition. I see it a lot here but it used to be more pervasive.

jae has done a good job over the years job at repeatedly beating people over the head with the same few arguments that have effectively rid this board of a few popular misnomers. Most significantly: " high points per game = good offensive player".

People don’t like to have to rethink things that they’ve taken as fact for, in some cases, their whole lives. It’s uncomfortable and it’s hard work to rethink a premise from which you’ve made other decisions. Now you have to think about all the other things you have to rethink. Sometimes it’s easier to be stubborn, plug your ears, shut your eyes and say things like “20 game winner, nuff said”.

I give pressure the reach-around.

by bloodsweatndonuts on Mar 18, 2010 11:31 AM PDT up reply actions

That’s a completely and utterly BS way to evaluate something. "I have an opinion based on no evidence, but someone taking the time to do research and come up with evidence is doing it wrong because it doesn’t agree with my opinion". We need to ditch this mentality. It’s stupid. It doesn’t make sense.

That is not my mentality and you have no idea how I evaluated their defense so how can you call it BS? All I did was state my opinion, not how I formulated it.

duaneok66’s numbers are no more valid than my opinion because that is all they are, his opinion. Nothing scientific about it. I watch the games very closely and make note of all the things he claims to keep track of. On Monday I was paying extra close attention to Monta’s defense on Kobe because I was interested to see how he did.

I feel like there are a lot of positives missing from that list and some of the negatives were a result of bad officiating.

Did Kobe score? Did it help the Lakers and hurt the Warriors? You mean, someone is actually being rated negative for a play that negatively affected our team? I don’t see what the objection is.

The objection is that Monta did everything right on that play. He didn’t play “bad” defense and thus shouldn’t be penalized. If someone is going to be “rated negative” it should be the official that was standing right there.

Monta played terrific defense on Monday. CJ was not the best defender on the floor for the Warriors. The opinion of some random dude in Oklahoma, who for all I know has no more basketball knowledge than me, is not going to change my mind . Plus, Jim Barnett, who I KNOW has more basketball knowledge than me AND duaneok66, seems to think Monta did a great job.

by WheresMyChippy on Mar 18, 2010 12:10 PM PDT up reply actions

Re: Nothing scientific about it.

It’s not objectively airtight, but it is recording data and applying the scientific method, which makes it scientific. I mean, “nutritional science” often tells us that we need to eat more of X and less of Y, and, over time, a lot of those end up being proven incorrect. It doesn’t mean the didn’t use the scientific method to come up with their conclusion, it just means that they came to the wrong conclusion. They may have set up their experiment wrong, collected the wrong data or did not include an important data point.

I watch the games very closely and make note of all the things he claims to keep track of. On Monday I was paying extra close attention to Monta’s defense on Kobe because I was interested to see how he did.

Right, but he’s writing it down (entering it into a spreadsheet) so it’s more accurate. No matter how good your memory is, it’s not as good as a spreadsheet’s (unless you forget to save).

I’m not saying that I agree with all of his metrics, but he doesn’t seem to be claiming that they are perfect or final. I think the more productive discussion would be to throw your ideas as to what should be observed and recorded as opposed to just disagreeing with his conclusions

To me, this is productive:

The objection is that Monta did everything right on that play. He didn’t play "bad" defense and thus shouldn’t be penalized. If someone is going to be "rated negative" it should be the official that was standing right there.

Regardless, I hope you end up finding your Chippy.

I give pressure the reach-around.

by bloodsweatndonuts on Mar 18, 2010 12:24 PM PDT up reply actions

Look, I’m all for good discussion, if you’ve taken the time to evaluate it and have real evidence behind your opinions, by all means, present it. I was so hostile initially because you’ve presented no evidence and are cherry picking one situation where your opinion doesn’t agree with what he’s done, but he’s actually presented the research behind it. It may not be correct, but if you’re going to argue against it, you have to provide real arguments – have you tracked how Monta did? Can you tell us how many turnovers he created and how many points he let up (or didn’t let up)? As bloodsweatndonuts pointed out – his method IS scientific, what evidence do you have that doesn’t agree with it? You didn’t give any, and I’m extremely hostile to the way it was presented, I see those kinds of arguments all the time (whether it be sports, politics, or anything, really), and that’s not a proper way to question research. One weird result doesn’t invalidate an entire systematic method of evaluation, and you have to show why the result doesn’t make sense to begin with, something you haven’t done so far.

The objection is that Monta did everything right on that play. He didn’t play "bad" defense and thus shouldn’t be penalized. If someone is going to be "rated negative" it should be the official that was standing right there.

I really can’t say I see a problem with a system that penalizes his for “doing everything right”, when the end result hurt his team. This is what sample size is for, if he’s really doing everything right, it will show up over time. No need to get up in arms about a little random variation in results over one game.

Plus, Jim Barnett, who I KNOW has more basketball knowledge than me AND duaneok66, seems to think Monta did a great job.

Maybe true, but he’s also a Warriors employee who’s job it is to talk up the team and be positive….

by Missing Barry on Mar 18, 2010 12:37 PM PDT up reply actions

Just to add a few more hits to the long dead horse, there are definitely some arguments that can be made against some of this stat. But the author has already claimed many of the stats, both + and -, are subjective. I imagine 10 different people analyzing the game using the same rules/stats will probably come up with 10 different scores.

The concept of trying to account for more defensive factors is good. I’m very interested to see where this stat goes once it is refined. See if it eventually can match up with considered “good defensive players”.

by FitzPRSpin on Mar 18, 2010 12:48 PM PDT up reply actions

See if it eventually can match up with considered "good defensive players".

I don’t think this should be the goal. IMO a lot of the defensive reputations around the league, whether they are good or bad, aren’t well deserved. I think we tend to let aesthetics play far too big a role when judging how good a player is at defense.

"I could be chasing an untamed ornithoid without cause."

by olympicmike on Mar 18, 2010 1:53 PM PDT up reply actions

To both of you

Basically, the problems I have with the system are addressed very well in bloodsweatndonuts’ post below this. I don’t like how over simplified it is. For it to truly work, each instance recorded must be revisited with the Step 2 and Step 3 he mentioned.

I guess, technically, it IS scientific. But there was a lot of “science” that said cigarettes aren’t bad for you. With how it is now, you would get vastly different results depending on who is filling out the spreadsheet. And as of now, only one person is doing that. I remember Monta being responsible for AT LEAST 12 of the turnovers, Kev has him at only 8. If there were a hundred people watching the game who filled out Kev’s spreadsheet and then they were all averaged that would be better. But still not adequate without Step 2 and 3.

Every shot contest is different, every steal is different. The same points can’t always be awarded for such a broad thing as “contested shot.” If the system were more fine tuned, and looked at each instance more closely, I’m pretty sure Monta would wind up above CJ on the list.

I wasn’t trying to cherry-pick. I was just bringing up an example of an instance were a player could be docked even though he played the right way. With a Step 2, this would not be a problem.

Once again, this is all nothing more than my opinion. I saw CJ at the top and Monta at the bottom and expressed what anybody who watched the game should have seen (in my opinion). Did you watch the game? Do you think CJ played better defensively than Monta? Did you think it before you looked at this post?

P.S. Jim Barnett is fo’ real tho. Dude doesn’t BS.

by WheresMyChippy on Mar 18, 2010 1:08 PM PDT up reply actions

To be fair to the author, he hasn’t elaborated on the finer points of his system, so he might have done a lot of the stuff we’d be looking for in a good system. Also, sample size solves a lot of the problems you’re bringing up.

I was just bringing up an example of an instance were a player could be docked even though he played the right way.

I’m not gonna say this is unfair, but everyone should realize this is a preference issue rather than a right/wrong issue.

expressed what anybody who watched the game should have seen (in my opinion). Did you watch the game? Do you think CJ played better defensively than Monta?

I wasn’t really making note of CJ’s defense, but I came away from the game wondering how effective Monta’s defense really was. How much is each forced turnover worth? How many times did Kobe score easily? Monta’s defense looked like a whole lot of flash over substance to me – he didn’t do a good job maintaining position or playing fundamental defense in general, but he did force a lot of turnovers. I can see how some people would look at the turnovers and think “great game”, especially with Fitz and Barnett highlighting every time Monta does it and talking it up. Were the turnovers enough to offset the otherwise poor performance of Monta in this case? I don’t know – I don’t know how many times each happened or exactly how much each event is “worth”, but it seems to me you’d need to know those things to make a quality evaluation. This is exactly what the author is trying to do. His methodology may or may not be accurate, but he’s presenting us with the exact type of information you need to make a good evaluation. Watching the game and relying on our memories is the exact opposite – we (people in general) are notoriously bad at objectively and accurately remembering things, throw in some announcers highlighting certain aspects and not others, and us rooting for one of the teams, and….well…..that’s not a good recipe. That’s why I ask for evidence.

As for Barnett, anyone selling fans on the Warriors has to BS. The product is terrible, and when you’re selling a terrible product, well….BS is your only choice. I do like Barnett, but he doesn’t represent anything close to “objectivity”.

Also, there might be an error in interpretation here. It doesn’t seem to me like the data for one game is telling us who played better. It’s telling us who’s defensive results were better. CJ was guarding Derek Fisher. Monta was guarding Kobe. Monta’s results should be worse than Kobe’s. This system presents each defensive event and scores it based on outcomes – all it’s trying to say it whose outcomes were better or worse.

by Missing Barry on Mar 18, 2010 1:50 PM PDT up reply actions

There it is

It’s telling us who’s defensive results were better. CJ was guarding Derek Fisher. Monta was guarding Kobe. Monta’s results should be worse than ~~Kobe’s~~CJ’s. This system presents each defensive event and scores it based on outcomes – all it’s trying to say it whose outcomes were better or worse.

I guess that’s my problem with the system. I was wanting it to do something it wasn’t meant to do: measure a players defensive abilities in comparison to others.

With a little tweaking though, this system could do that. Score based on actions not outcomes. Example: 1) Player A has a hand in Kobe’s face and he still hits a fade away. 2) Player B gets there a little late but Farmar misses a jumper. If we want to know who played better defense the better score of the two should go to player A.

All in all I just wanted to refute the claim that CJ played better defense than Monta. I now see that claim was not being made.

And I don’t know, I think Jim is pretty darn objective. He calls the Warriors out when they do things wrong and he praises the other team when they do things right. But then again he is sitting next to Fitz, whose bias is so strong that it creates a gravitational field that warps our perceptions of Jim, making him look more objective by comparison.

by WheresMyChippy on Mar 18, 2010 3:22 PM PDT up reply actions

Yeah I can’t stand Fitz. He’s ridiculous.

With a little tweaking though, this system could do that.

I do agree – with what you’re looking for, the process is more important than the results. The one problem I see is I think you might lose a lot of hard to define gray area that matters when you don’t look at results, like two guys might challenge shots an equal number of times, but one guy does it better and causes his player to miss more often. With the results based approach, with a large enough sample size it should start to tell you the same thing – in any given game the results mess things up, but over time, if you are challenging shots, guys will miss more often and you’ll rack up a better defensive score.

So yeah, there’s probably a lot more that can go into it, but I think it’s a great start in the right direction.

by Missing Barry on Mar 18, 2010 5:54 PM PDT up reply actions

P.S. Jim Barnett is fo’ real tho. Dude doesn’t BS.

but he’s too close with the players to always be honest about them. He don’t like to hurt their feelings or their appeal to the fans so he often just says the polite thing and leaves the criticism out.

Lights please, lights please, turn off the lights.

by Skeptic con Urquell on Mar 18, 2010 9:43 PM PDT up reply actions

P.S. Jim Barnett is fo’ real tho. Dude doesn’t BS.

I think he’s, if nothing else, sincere. You’re right that he’s not going to go out of his way to criticize the team, but he does it more often than any other announcer being paid by a team.

I think that when he does say something, it is objective. The only homerism might come from not mentioning something a Warrior’s player had done wrong that he might have if another team’s player had. But that’s the extent of it.

He does a really good job at being objective, clear and detailed. He also seems like a really nice guy. unlike Vlad who seemed like kind of a d*ck on the broadcast.

I give pressure the reach-around.

by bloodsweatndonuts on Mar 18, 2010 10:13 PM PDT up reply actions

I think that when he does say something, it is objective. The only homerism might come from not mentioning something a Warrior’s player had done wrong that he might have if another team’s player had. But that’s the extent of it.

Oh no, I love Jim and his broadcast’s, I’m just saying as a player he won’t say things about them that others might, and not just the dubs players, Jim likes them all equally. and If I was the announcer I’d do it the same way that Barnett does it , cause I’m sweet like that. on second thought I’d probably strangle Fitz but other than that I’d be Barnett.

Lights please, lights please, turn off the lights.

by Skeptic con Urquell on Mar 18, 2010 10:32 PM PDT up reply actions

Yep, announcers are paid to be positive. Nobody would listen to a negative announcer. Trying to argue that Barnett is “objective” is a bit ridiculous. He might be one of the more fair announcers out there, maybe even the most fair, but no announcer is objective, it isn’t their job to be objective – their job is to give you an enjoyable game experience and build up as much interest as possible so as many people as possible watch.

Again, just think about it this way – the Warriors are a garbage organization from the top to the bottom. Can any announcer really be positive and objective at the same time while talking about the Warriors? Absolutely not. If you want, listen closely – not listening for evidence to back up your point, but rather with an open mind that maybe he is extremely positive overall. You might be surprised by what you find. Giants announcers are the same way – I love Kruk and Kuip, and I think they’re less homers than pretty much any other announcer I’ve heard do baseball, but they’re still Giants employees and nowhere close to objective.

by Missing Barry on Mar 19, 2010 6:58 AM PDT up reply actions

+1

I watch alot of games on NBA league pass. They generally give you the home teams feed and the most objective color guy (imo of course) is suprisingly Sean Elliot of the Spurs broadcasting team. Barnett is good and Fitz is a tool.

Thing B

by warriorsscore110 on Mar 19, 2010 1:19 PM PDT up reply actions

How was Vlad a jerk on the broadcast?
He was talking about a foundation for helping homeless children. Maybe you have no soft spot in your heart for people who do good things like that for others. Maybe hes not the greatest broadcaster, but english is his second language. i thought personally he was pretty humble, when Fitz said he is a great shooter, and he interrupted him and said he had been terrible this year, and etc.
Curious, what did he do to make you feel that way?

Nellie needs to go! but while we suffer....John Wall or Evan Turner please!

by MDB on Mar 19, 2010 5:17 PM PDT up reply actions

Re: How was Vlad a jerk on the broadcast?

I thought he was a bit disrespectful towards Jim, bagging on him. Taking advantage of the fact that Jim really likes to be considered one of the guys so he’ll take some flack. I thought that was kind of a d*ck move. Show some respect. Plus his jokes weren’t really funny. Just rude.

I’m not asserting this as fact, just saying that’s how I took it. I also I’m also a huge Jim Barnett fan and Vlad has sucked as a Warrior so I don’t feel like he’s earned the right to cap on someone I respect. That’s just my opinion, so take my comment with a grain of salt. I’m not trying to convince anyone of my point-of-view. I was just wondering if anyone else felt that way.

I give pressure the reach-around.

by bloodsweatndonuts on Mar 19, 2010 8:10 PM PDT up reply actions

Ok, i didn’t see it the same way as you, and the joke he made was pretty harmless it seemed, and they all had a little chuckle.

Nellie needs to go! but while we suffer....John Wall or Evan Turner please!

by MDB on Mar 19, 2010 8:22 PM PDT up reply actions

Some random feedback by someone who is poor at statistics

This is great. I hope people don’t get too hung up on the scoring or the semantics of statements like “bad defense”. They should be taken in the context of your experiment and should not require you to qualify every statement like “bad defense according to my still developing system”.

I like this defensive charting because you’re isolating events and recording them. That’s step 1 in understanding something like this. I’m not too concerned with the scores that it’s returning at this point. I’m more interested in the amount of times these events are happening and thinking about the validity and impact of each of those events on the opposing offense.

I would guess, step 2 would be revisiting all of the events being recorded and adding, subtracting or adjusting some. Step 3 might be weighting them for impact.

I would also like to see every event have a natural inverse. Steals versus attempted steal resulting in an advantage for the offense. If you’re saying some action is hindering the offense, there must be the opposite of that action (even if it is specific inaction) that helps the offense. The inverses wouldn’t be equally weighted though based on the level of difficulty (like blowing up a transition break when outnumbered versus Lazy Transition D).

Examples:

Points are awarded for:

     * Blocks / Attempt at block taking you out of defensive or rebounding position
    * Steals / Attempt at steal resulting in loss of defensive position or a foul
    * Loose Balls / Taut scrotum?
    * Contesting a shot (that is missed) / You have this
    * Forcing a turnover (knocking a ball off an opponent’s leg, for example) / attempt at doing this resulting in loss of defensive position or a foul
    * Drawing a charge / Blocking foul or a flop resulting in loss of defensive position
    * Hedging (forcing a guard to stop or reverse direction on a pick) / Losing your man allowing him to roll into scoring position
    * Blowing up a transition bucket when outnumbered / Foul resulting in 3-point-play opportunity (inverse could be fouling a <70% ft shooter) and you have lazy transition D

Points are deducted for:

    * Giving up penetration / denying without fouling
    * Failure to contest
    * Going under screens on a three point shooter / fighting through without giving up the drive
    * Lazy transition defense
    * Trailing on screens / playing this properly depending on the offensive player
    * Shooting Foul / this is the inverse of a lot of the positives like contesting, or steals, or forced TO.
    * Failing to Box Out (and giving up an offensive rebound) / good box out
    * Getting Posted up and giving up a score/ not allowing the post player to get to the paint or forcing a fade-away
    * Getting Posted up and needing a double team / Not needing a double team

One other thing, and I know you probably disagree: My opinion is that the result should be removed from things like contested shots. I’d assume that eventually you’d want to be able to predict repeatability. If a defender contests a shot 10 times, it shouldn’t matter how many times that shot went in. The event of contesting has less variability than the event of a contested shot that misses. I would even argue that contested fg% from different zones should be a metric on the offensive player. Then you could try and predict outcomes based on who is defending who.

Regardless, this is awesome. Thanks for posting it!

I give pressure the reach-around.

by bloodsweatndonuts on Mar 18, 2010 12:04 PM PDT reply actions

Thank You

Step 2 and Step 3 are CRUCIAL. I feel this system doesn’t truly work as a good evaluation of defense without them.

by WheresMyChippy on Mar 18, 2010 12:14 PM PDT up reply actions

Personally, I like using the result. I think there are aspects that are hard to quantify – things like “how much did he contest the shot”, as surely all contesting of shots isn’t equal in how it affects shooting percentage, and taking into account the results will capture a lot of little things like that. It would probably raise sample size requirements, though.

by Missing Barry on Mar 18, 2010 12:41 PM PDT up reply actions

re: It would probably raise sample size requirements, though.

Right. The NBA or some company (assuming teams would buy this data) should watching every game and recording this stuff.

I think there are aspects that are hard to quantify – things like "how much did he contest the shot", as surely all contesting of shots isn’t equal in how it affects shooting percentage

Agreed but you’ll never dial in anything 100% right? That’s why I was advocating for combining contested shots with the offensive player’s contested fg% by zone. Then based on how often an offensive player shots form a given zone you can try and predict how successful he will be against a given defender, in part by, how often that defender contests shots.

But you’re dead on about the variation in the strength of a contested shot. I’m just thinking about how to eliminate as much noise as possible.

I give pressure the reach-around.

by bloodsweatndonuts on Mar 18, 2010 12:57 PM PDT up reply actions

Yep, it definitely seems to me like a balance between noise and trying to capture unseen variables/skills. Not sure there’s necessarily a right answer.

by Missing Barry on Mar 18, 2010 1:00 PM PDT up reply actions

I’m just thinking about how to eliminate as much noise as possible.

The easy way is start with the league standings, then look at the players on each winning team, and you’ll get 99% of the info needed to rank them with little need to make watching the games a counting and recording exercise? Game flow doesn’t stop to be counted or paused to be digested then resume in real time, it’s a product of the moment, now, then gone.
I really question how one guy can watch 10 players on a court and rate how they are all interacting without knowing all the conditions such as floor condition, ball, shoes, what they all ate before the game, etc. What you think you see is not always what is really happening, like did that guy just now go at Montay because he was afraid to go at hunter,?so hunter should have really got the stop??

Lights please, lights please, turn off the lights.

by Skeptic con Urquell on Mar 18, 2010 9:58 PM PDT up reply actions

So 50/50

I know better then to go down the rabbit hole with my dear old dad.

Let’s just say, to differing extents, I agree with the second paragraph and disagree with the first.

I give pressure the reach-around.

by bloodsweatndonuts on Mar 18, 2010 10:29 PM PDT up reply actions

I agree with the second paragraph and disagree with the first.

that’s more than I ever hoped for.

Lights please, lights please, turn off the lights.

by Skeptic con Urquell on Mar 18, 2010 10:33 PM PDT up reply actions

I disagree with almost all of it. But not quite all. ;)

by Missing Barry on Mar 19, 2010 6:59 AM PDT up reply actions

One thing to note...

Like others have said, you can’t throw out a metric because Monta scored poorly against a great player while giving a solid effort. Look, I would worry more about this system if Monta got a good score playing against someone like Kobe. Guys like Kobe make defenders look bad, despite giving their best effort.

If you allow for a large enough sample the Kobe’s of the world and the Courtney Lee’s of the world would cancel each other out. Monta might score worse than his defensive ability against the great players, but he’ll score better than his ability against some of the lesser players that he will match up with over the course of a season.

Overall I think the system is a pretty solid base. I’m sure it could use some tweaking, but it is a great place to start. Putting it up for all to see on forums like this one is a great way to set about the refining process as there are some pretty sharp basketball minds floating around. Keep up the good work.

"I could be chasing an untamed ornithoid without cause."

by olympicmike on Mar 18, 2010 1:09 PM PDT reply actions

Exactly

If you take any offensive player and track their performance using any metric ever created, I guarantee you they’ll score much worse if they’re being guarded by Dikembe Mutumbo in his prime than they would being guarded by Troy Murphy.

If the Warriors played 20 games against the Lakers, they’d probably go 2-18 and look pretty bad in a vacuum. If they played 20 games against the Nets, they’d probably go 15-5 and look pretty good in a vacuum.

What’s my point? You cannot use any measure of anything in a vacuum. So stop trying. Every measure of everything will never perfectly account for extenuating circumstances. This is not Pi. You cannot boil anything down to a number.

Trust me, learning english isn’t a waste of time. It is actually sort of useful.

-randolphforpresident

by Dubs fan in Boston on Mar 18, 2010 4:18 PM PDT up reply actions